108 Alpha: Retirement Bug

[Pang Bingxun]

The retirement bug ist still alive. If you increase the manpower gain by agro techs, ministers and probably also the defence lobby slider you increase the rate of retirement. Therefore the average duration of army service depends on all those variables by a factor of up to 2.39 if i calculated that correctly.

 

[Radu]

Right and wrong at the same time.

Wrong, because manpower growth has no effect on aging directly. It's not a factor in the formula, simple as that.

Right, because if manpower growth outpaces aging and reinforcement drain, the manpower pool increases, of course. And the larger the manpower pool, the larger the aging manpower value.

Also, aging is calculated as ( Manpower Pool + Manpower in Built Military Units + Manpower in Buildings on Map ) * Aging Factor.

So the more things you have on the map, the more aging you will experience.

It is likely the aging due to manpower in military units and buildings that throws you off.

Many things happen all at once during regular gameplay of a full-scale campaign, that's why it's sometimes a good idea to test conclusions in the "controlled environment" of a tiny test battle-scenario.

 

[Pang Bingxun]

Wrong, because manpower growth has no effect on aging directly. It's not a factor in the formula, simple as that.

Are you sure about this? I am aware that manpower growth will increase retirement in that way that there are more people that can retire. But that is not what i mean. The retirement can change within a day if a minister is changed. Here an example:

Playing china i have a total manpower of 13613.7. That should be enough to deal with "rounding errors". By Agrotechs i have a multiplicator of 1.2. By full defence lobby i have another multiplicator of 1.1. I also have the ministers man of the people(+10%) and school of mass combat(+25%). So there is a total multiplicator of 1.2 x 1.1 x (100%+10%) x (100%+25%)= 1.815. Am i right so far?
There retirement shown in the mouse over at the manpower symbol says i have a daily retirement of 2.47. Now i change the chief of staff giving me a total multiplicator of only 1.2 x 1.1 x (100%+10%) = 1.452. Now the daily retirement shown in the mouse over has shrunk to 1.98 and the total manpower is still 13613.7. So the retirement shown in the mouseover seem to be calculated by (total manpower) x (tech modifier) x (defence lobby modifier) x (security minister modifier) x (chief of staff minister modifier) x (national identity modifier) x (social policy modifier) x (national culture modifier) x 0.0001.

There is a slight possibility that the retirement shown is not the real retirement, but i don't think so.

 

[Radu]

Are you sure about this? I am aware that manpower growth will increase retirement in that way that there are more people that can retire. But that is not what i mean. The retirement can change within a day if a minister is changed. Here an example:

Playing china i have a total manpower of 13613.7. That should be enough to deal with "rounding errors". By Agrotechs i have a multiplicator of 1.2. By full defence lobby i have another multiplicator of 1.1. I also have the ministers man of the people(+10%) and school of mass combat(+25%). So there is a total multiplicator of 1.2 x 1.1 x (100%+10%) x (100%+25%)= 1.815. Am i right so far?
There retirement shown in the mouse over at the manpower symbol says i have a daily retirement of 2.47. Now i change the chief of staff giving me a total multiplicator of only 1.2 x 1.1 x (100%+10%) = 1.452. Now the daily retirement shown in the mouse over has shrunk to 1.98 and the total manpower is still 13613.7. So the retirement shown in the mouseover seem to be calculated by (total manpower) x (tech modifier) x (defence lobby modifier) x (security minister modifier) x (chief of staff minister modifier) x (national identity modifier) x (social policy modifier) x (national culture modifier) x 0.0001.

There is a slight possibility that the retirement shown is not the real retirement, but i don't think so.

*sigh* ... let's review :

1) Aging is a factor applied to all of the manpower in the realm. Its value is a fraction of the total manpower.

That means if you have 100 manpower everywhere ( pool,military,buildings ) and zero growth and zero drains and the aging factor is 0.01 (1%) , you will lose 1 manpower to aging.

2) Manpower growth increases your manpower ( if it outpaces reinforcements drain ).

That means that if your growth is 10 manpower (and no other drains), your manpower base is 100 + 10.

Aging is applied to 110. That means 1,1 manpower lost to aging.

But what if your manpower growth is 20?

Then of course the aging base is 100 + 20 and that results to manpower lost to aging being 1,2.

Again, where is the issue? Yes, manpower growth increases your manpower ( I can't believe I'm typing such an obvious truism ).

Bigger growth => more manpower => bigger aging.

Really, isn't it elementary maths that if you divide a bigger number (more manpower due to more growth) by the same ratio, you get a bigger remainder ( more manpower lost to aging in absolute value )?

 

[Pang Bingxun]

This is not the issue. It is quite obvios that changing manpower will change retirement, although this usually takes quite some time.

The issue is that even if manpower does not change retirement can change within a day by a factor of almost 2. When i was speaking of total manpower i meant total manpower or "Arbeitskräfte komplett" in the german translation. It is the sum of unused manpower, manpower in the military, manpower in provinces/buildings, manpower in current production.
The problem is that the same number is divided by different factors.

 

[Radu]

This is not the issue. It is quite obvios that changing manpower will change retirement, although this usually takes quite some time.

The issue is that even if manpower does not change retirement can change within a day by a factor of almost 2. When i was speaking of total manpower i meant total manpower or "Arbeitskräfte komplett" in the german translation. It is the sum of unused manpower, manpower in the military, manpower in provinces/buildings, manpower in current production.
The problem is that the same number is divided by different factors.

But that's the point. In your case manpower does change .

(simplified)
Your Manpower Base = Manpower Pool (unchanged) + Projected Manpower Gain (you change it!)

You Manpower Base changes , thus your Aging Manpower also changes.

And its behaviour is completely WAD to both the original intent of the feature and the internal maths.

 

[Pang Bingxun]

(simplified)
Your Manpower Base = Manpower Pool (unchanged) + Projected Manpower Gain (you change it!)

So if Manpower Base = 13613.7 + 2.92 changes to Manpower Base = 13613.7 + 2.34 this will change the retirement by 20% from 2.47 to 1.98. Is this the way it was intended?

PS: 2.92 and 2.34 are the Projected Manpower Gains per day. All those numbers are from a save i just opened.

 

[Radu]

So if Manpower Base = 13613.7 + 2.92 changes to Manpower Base = 13613.7 + 2.34 this will change the retirement by 20% from 2.47 to 1.98. Is this the way it was intended?

PS: 2.92 and 2.34 are the Projected Manpower Gains per day. All those numbers are from a save i just opened.

That's correct.

 

[Denniss]

I'm really glad we disabled this feature for CORE - makes minister boni almost useless with this big jump in retirement.

It would be better if retirement is applied to total manpower (in pool or used in military or buldings) only, without adding any other effects.

 

[Pang Bingxun]

makes minister boni almost useless with this big jump in retirement.

In the extreme situation of maximum manpower minister boni will even have no effect. I define maximum manpower by (daily manpower gain) = (daily retirement). Then minister boni will apply on both sides of the equation:

(minister boni) x [(daily manpower gain) - (daily retirement)]= 0.

Communist China has even the really strange situation of (daily manpower gain) < (daily retirement). Therefore ministers that should increase manpower will decrease it instead. Therefore i prefer to think about the feature retirement in the current implementation as a bug. It enforces to waste manpower excessively instead of losing it anyway due to retirement. A less aggressive implementation would be preferable.

Currently aging is calculated as ( Manpower Pool + Manpower in Built Military Units + Manpower in Buildings on Map + Manpower in current Production) * Aging Factor. If this Aging Factor has a constant value, than retirement would work fine and manpower ministers will so, too.